In the following section the result and their subsequent analysis of relevant clinical and other investigations features have been detailed. 80 cases of lean type 2 diabetes mellitus, 80 cases of non obese type 2 diabetes mellitus (BMI >20.25-<25) and 80 cases of obese type 2 diabetes mellitus (BMI>25) were selected for present study. Table 1,Table 2,Table 3,Table 4,Table 5,Table 6,Table 7,Table 8

Among lean cases 65% are male and 35% are female. Among non obese 70 % are male and 30 % are female. Among obese 45 % are male and 55 % cases are female.

Among lean subjects mean duration of diabetes, mean fasting and post prandial blood glucose and glycosylated hb values were 5.5 years, 226.5 mg/dl, 282.5 mg/dl, and 9.15 respectively. Among obese these values were 5.75 years, 193.1 mg/dl, 248.65 mg/dl and 7.81 respectively. While among obese these were 6.1 years, 206.65 mg/dl, 262.7 mg/dl and 8.44 respectively.

Mean values of total cholesterol and LDL cholesterol were highest in obese subjects. While mean values of VLDL cholesterol, HDL cholesterol and triglycerides were higher in lean subjects.

Table 4

Data	lean		Non obese		obese
	No.	%	No.	%	No.	%
Sugar	80	100	72	90	76	95
Microproteinuria	12	15	12	15	12	15
Overt proteinuria	4	5	8	10	16	20
Pus cells	4	5	-		4	5
rbc	-		2	5	-

Sugar in urine was present in 100 % of lean patients and 90 % and 95 % in non obese and obese patients respectively. Microprotienuria was present in 15 % of each group.

Mean values of urea and creatinine were highest in lean subjects among all the three groups.

In this analysis HDL-c, VLDL-C, urea and creatinine were not found to be significant. So next analysis was done with those dependent variables which showed significance.

Table 7

Dependent variables	weight type (i)	Weight type(j)	Mean difference(i-j)	Std. error	significance	Remark
FBG	lean	Non-obese	33.4	8.18	0.001	significant
PPBG	lean	Non-obese	34.2	9.70	0.001	significant
TC	lean	Non-obese	-19.6	6.38	0.003	significant
LDL-C	lean	Non-obese	-26.5	6.77	0.001	significant
TG	lean	Non-obese	15.2	6.75	0.028	significant

Table 8

Dependent variables	weight type (i)	weight type (j)	Mean difference(i-j)	Std. error	signifiacnce	Remark
FBG	lean	Obese	19.8	8.18	0.018	significant
PPBG	lean	Obese	20.1	9.70	0.042	significant
TC	lean	Obese	-21.3	6.38	0.001	significant
LDL-C	lean	Obese	-27.8	6.77	0.001	significant
TG	lean	Obese	17.6	6.75.	0.011	significant

Type 2 diabetes mellitus accounts for the 85% of the people with diabetes worldwide. Development of the disease is summation of environmental insult to genetically predetermined metabolic disruption. Type 2 diabetes mellitus is characterized by the pathophysiologic abnormalities, impaired insulin secretion, peripheral insulin resistance and excessive hepatic glucose production. Obesity visceral or central is very common in type 2 diabetes mellitus. Low body weight type 2 diabetes mellitus, phenotypically a separate type of type 2 diabetes mellitus is of interest in tropical region. Characterized by its typical age presentation, altered lipid pattern along with some controversial renal involvement, more elevated fasting and post prandial blood glucose at the time of presentation of low body weight type 2 diabetes mellitus in comparison to obese and non obese type distinguishes it further. The increased mortality among lean diabetic smokers has been observed in the general population as well6,7,8,9. A possible explanation for the observed obesity paradox could be sarcopenic obesity, defined as the presence of high body fat with reduced or normal lean body mass10. Sarcopenic obesity reduces the cardiopulmonary fitness and physical functioning possibly leading to premature death and could account for the higher mortality eventually seen in individuals who are normal weight at the time of onset of diabetes mellitus11.

C.S. Yagnik et al depicted in his study12 the male preponderance in lean type 2 diabetes mellitus, which is comparative with the sexual distribution in present study which showed 65 %male and 35% female distribution among lean type 2 diabetes mellitus.

Body mass index is the single most important predictor of low body weight type 2 diabetes mellitus and other phenotypically different types of type 2 diabetes mellitus. According to Indian authorities and WHO if the body mass index is less than 20.25, type 2 diabetes mellitus is considered as lean body weight. In the present study 17.91 is the mean BMI of type 2 diabetes mellitus belonging to low body group. This is at par with the observation of other and was significantly lower than non obese (23.01) and obese (27.44).

One important observation is that fasting blood glucose in low body type 2 diabetes mellitus is higher (226.5 mg/dl) in comparison to non obese (193.1 mg/dl) and obese ( 206.65 mg/dl). It is consistent with previous large series of observation by S. Das et al13, B.K. Sahay and K.Kannan, V. Seshaish et al14 and Sameer Banerjee. Post prandial blood glucose was also higher in case of lean type 2 diabetes mellitus as compared to non obese and obese ( 282.85 mg/dl vs 248.65 mg/dl vs 262.7 mg/dl ). These observation are associated with significantly high mean glycosylated hemoglobin level in lean type 2 diabetes mellitus than non obese and obese cases (9.15 vs 7.81 vs 8.44). This data is consistent bwith studies of S Das and V Sidharta et al15 .In this study using general linear model multivariate with controlling for age and POST HOC LSD, FBG, PPBG and glycosylated hemoglobin was found to be significantly increased in lean compared to non obese and obese. All the above observation suggest that low body weight type 2 patients are having less insulin secretion either4 in fasting state or post prandial state and it is also evident that day to day regulation of insulin secretion is also defective.

Hypercholesterolemia i.e. cholesterol level >240 mg/dl (10%) in contrast to hypertriglyceridemia i.e. triglyceride level >200 mg/dl (35%) was less common in low body weight type 2 diabetes mellitus but HDL cholesterol value was not significantly variable in the low body weight as compared to non obese and obese. Lowe incidence of hypercholesterolemia and relatively higher incidence of hypertriglyceridemia in low body weight was also found in studies of K. Kannan, C.S. Yagnik et al12.

The increased incidence of raised blood urea and serum creatinine was found more in low body weight as compared to other two groups. It is consistent with finding by Sidhartha Das. But this difference was not significant.

The most characteristics criteria in low body weight type 2 diabetes mellitus i.e. lower body mass index has got increased fasting and post prandial blood sugar values in comparison with non obese and obese patients which reflect decreased beta cell reserve having lesser amount of endogenous insulin secretion. Hypertriglyceridemia, reflecting lower body insulin status is characteristics of low body weight type 2 diabetes and all these findings came out to be significant in our study also.

Hoet J J Tripathy B B Report of the international workshop on types of diabetes peculiar to the tropics Diabetes care 1996 19 1014 1014 Tripathy B B Kar B C Observations and clinical patterns of diabetes in India Diabetes 1965 14 404 416 Samal K C Das S Agarwal B N Panda N C Tripathy B B Nutritional status and profile of NIDDM of recent onset J Diab Assoc India 1998 28 99 101 4. Ramchandran A Snehalatha C Latha E Vijay V Vishwanathan M Rising prevalence of NIDDM in urban population in India Diabetologia 1997 40 232 239 Mohan V Vijayaprabha R Rema M Premalatha G Poongothai S Deepa R clinical profile of lean NIDDM in south India Diabetes Res Clin Pract 1997 38 101 109 Zheng W Mclerran D F Rolland B Zhang X Inouse M Matsuo K He J Gupta P C Ramdas K Tsugane S Association between body mass index and risk of death in more than 1 million Asians N Engl J Med 2011 363 719 748 Manson J E Willett W C Stamfer M J Colditz G A Hunter D J Hankinson S E Hennekens C H Speizer F E Body weight and mortality among women N Engl J Med 1995 333 677 85 8. Pishon T Boeing H Hoffemann K B Ergmann M Schulze M B Overvad K Van Der Schouw Y T Spencer E Moons K G Tjonneland A General and abdominal adiposity and risk of deathy in Europe N Engl J Med 2008 359 2105 120 Manson J E Stamfer M J Hennekens C H Willet W C Body weight and longevity. A Reassesment JAMA 1987 257 353 58 Floerzh Castillo-Floerz . S Beyond the obesity paradox in diabetes: fitness, fatness and mortality JAMA 2012 308 619 639 Zamboni M Maxzzali G Zoico E Harris T B Meigs J B Francesco Di Fantin V Bissoli F Bosello L . O HJealth consequences of obesity in the elderly: a review of four unresolved questions Int J Obes (Lond) 2005 29 1011 1040 Yagnik C S Aossiacition of obesity with clinical, biochemical, metabolic and endocrine measurement in newly diagnosed NIDDM patients- NovoNordisk Diabetes update. Kapur A Ed 1993 139 185 Das S Samuel S C Baliarsinha A K Tripathy B B Lean (underweight) NIDDM- Peculiarities and differences in metabolic and hormonal status- A pilot study J A ssc Physicians India 1995 43 339 381 Seshaiah Vet al: Lipid profile and other risk factors associated with vascular complications in NIDDM- Lipid India ILIB 1996 10 15 Das Siddharta: Lean NIDDM: An independent entity-in Novonordisk Diabetes updates proceedings Ed. Kapur A. Health Care Communication Bombay 1993 153 162